Continuous Ketone Monitoring for People with Diabetes: International Expert Recommendations for a New Technology

Introduction and Context

Diabetic ketoacidosis (DKA) remains one of the most serious acute complications of diabetes, carrying substantial morbidity, mortality, and healthcare costs. Although DKA is classic in type 1 diabetes, it also occurs in type 2 diabetes and has become more frequent in settings of euglycemic DKA associated with sodium–glucose co-transporter-2 (SGLT2) inhibitor therapy. Current international diagnostic frameworks for DKA rely on a triad: hyperglycaemia, metabolic acidosis, and ketonemia/ketonuria. Historically, ketone assessment has been episodic—urine dipstick or point-of-care (POC) blood beta-hydroxybutyrate (BHB)—measured during sick days or clinical review.

Continuous ketone monitoring (CKM) is an emerging technology analogous to continuous glucose monitoring (CGM). CKM sensors sample interstitial fluid and produce ongoing ketone concentration estimates, potentially alerting users to rising ketone burden hours before an episodic test would be done. Recognizing both the promise and the potential harms (alarm fatigue, anxiety, false reassurance), an international expert panel convened to create practical recommendations for CKM application. Their consensus, published by Dhatariya et al. and endorsed by the International Society for Pediatric and Adolescent Diabetes (ISPAD), aims to define actionable thresholds and workflows for clinicians and people with diabetes to use CKM safely and effectively (Dhatariya et al., Lancet Diabetes Endocrinol. 2026).

Why this guidance matters now: CKM could change how we prevent and detect DKA. But without agreed thresholds and actions, adoption risks inconsistent use, over-alarming, or missed opportunities to prevent acute deterioration. The expert consensus provides an early clinical framework while rigorous outcome trials and regulatory experience accumulate.

New Guideline Highlights

Major themes and takeaways from the expert consensus:

– CKM is a complementary tool, not a replacement for clinical judgement, blood pH measurement, or access to emergency care.
– The panel proposes pragmatic CKM alarm thresholds and stepwise action plans (escalation ladders) calibrated to limit alarm fatigue while promoting timely responses to clinically significant ketonaemia.
– CKM is particularly recommended for people at higher risk of DKA: those with type 1 diabetes on intensive insulin therapy, people with prior DKA, insulin pump users, pregnancy with diabetes, and persons using SGLT2 inhibitors.
– Special population adjustments are advised for children and pregnant women where physiologic considerations warrant lower thresholds or more conservative action.
– Education, sick-day planning, integration with insulin delivery/CGM data, and remote monitoring pathways are core implementation requirements.

Key clinical message: CKM provides dynamic ketone trajectories that, when paired with glucose and clinical context, can help prevent progression to DKA—but only if users and teams adopt standardized thresholds, clear actions, and supportive education.

Updated Recommendations and Key Changes

This expert consensus is novel rather than a revision of a previous CKM guideline; its principal contribution is to translate point-in-time ketone thresholds used for laboratory and POC testing into continuous-monitoring–friendly thresholds and actions. Important updates compared with traditional guidance include:

– From episodic to continuous: translating static BHB cut-offs (e.g., 3 mmol/L suggestive of DKA) into alarm bands and trend-based triggers for CKM.
– Trend emphasis: recommendations stress rate-of-rise and persistence (e.g., rising ketones over a defined time window) rather than single isolated values alone.
– Integration with glucose: explicit protocols for euglycemic presentations (e.g., SGLT2 inhibitor–associated DKA) where ketones may rise despite near-normal glucose.
– Pediatric and pregnancy-specific advice: endorsement by ISPAD supports child/adolescent-specific thresholds and heightened vigilance in pregnancy.

Evidence driving these adaptations is largely expert-driven consensus allied to existing DKA diagnosis literature (historical POC thresholds and clinical studies) and early device performance data. The panel emphasizes that their thresholds are provisional pending prospective outcome trials.

Topic-by-Topic Recommendations

Below are the core recommendations and practical action algorithms proposed by the expert group. Where possible, numeric thresholds are provided; these are consensus-based and intended for CKM devices reporting beta-hydroxybutyrate equivalents in mmol/L.

Who should be offered CKM?
– Priority groups: people with type 1 diabetes on intensive insulin therapy, those with a history of DKA, insulin pump/closed-loop system users, pregnant people with diabetes, individuals using SGLT2 inhibitors, and people with limited ability to recognize early sick-day symptoms.
– Consider offering CKM in people with recurrent unexplained hyperketonaemia, high HbA1c, social or geographic barriers to rapid healthcare access, or frequent intercurrent illness.

Recommended CKM alarm thresholds and actions (consensus framework)
– Normal/Green: <0.6 mmol/L
– Action: routine diabetes self-management. No immediate action for small, isolated excursions below 0.6.
– Caution/Amber: ≥0.6 to <1.5 mmol/L
– Action: confirm with a POC blood ketone test if symptomatic or unwell; review glucose and insulin delivery; initiate sick-day measures (oral fluids, maintain insulin dosing, carbohydrate intake as appropriate); repeat CKM trend monitoring. If ketones fall, continue close monitoring at home.
– Action/Red: ≥1.5 to 0.5 mmol/L per hour) or ketones rising despite corrective measures
– Action: treat as possible DKA—seek immediate medical evaluation. If accompanied by signs of acidosis (vomiting, severe abdominal pain, altered consciousness) go to emergency department for venous/arterial blood gas, electrolytes, and formal DKA management.

Notes on trends and context:
– Rate of increase: a rapid upward slope (e.g., >0.3–0.5 mmol/L per hour) is a red flag even when absolute values are modest.
– Symptom pairing: any alarm should be interpreted with concurrent symptoms (nausea, vomiting, polyuria, polydipsia, abdominal pain) and glucose data. A normal CKM reading does not exclude early DKA if acidosis is present; arterial/venous blood gas remains the diagnostic standard for acidosis.
– SGLT2 inhibitor users: maintain a lower threshold for clinical action and a lower barrier to in-person assessment because euglycemic DKA can occur with glucose 2 hours).

Expert Commentary and Insights

The panel’s key perspectives:
– Balance benefit and burden: Experts stressed that CKM’s success depends as much on human systems (education, skillful triage, and access to care) as on sensor performance.
– Avoid overmedicalization: Conservative alarm strategies and tiered actions were proposed to reduce alarm fatigue and anxiety; the panel favored trend-informed action rather than reacting to single borderline values.
– Evidence gaps acknowledged: The recommendations are consensus-based, reflecting the current absence of large randomized outcome trials showing CKM reduces DKA incidence. The panel called for prospective studies of CKM effectiveness, device accuracy across physiologic states, and health-economic analyses.
– Equity and access: CKM’s potential to reduce DKA may disproportionately benefit people facing barriers to rapid care—but only if devices are accessible; experts urged payers and health systems to consider equitable deployment strategies.

Areas of controversy
– Exact numeric thresholds: While the panel suggested thresholds (≥0.6, ≥1.5, ≥3.0 mmol/L), some clinicians argued for different cut points, particularly in pregnancy and younger children.
– Remote monitoring burden: There is debate over how many alerts should trigger clinician contact versus automated patient-directed actions; resource-constrained clinics may struggle with high alert volumes.

Practical Implications

For clinicians
– Prepare to incorporate CKM data into routine sick-day care pathways and triage protocols. Update clinic sick-day education and supply lists to include blood ketone testing and CKM interpretation.
– When prescribing SGLT2 inhibitors, counsel patients about euglycemic DKA risk and consider CKM for those at especially high risk or limited access to emergency care.
– Ensure clarity in documentation: who responds to CKM alerts, expected response times, and escalation steps.

For people with diabetes and caregivers
– CKM can provide early warning but requires an action plan. Learn your device’s alarm thresholds, how to confirm readings with a POC BHB test, when to call your diabetes team, and when to go to the emergency department.
– Keep supplies for POC BHB testing and a written sick-day plan readily available.

Health systems and payers
– Incorporate CKM into broader DKA prevention strategies, with attention to cost-effectiveness and equitable access. Pilot programs should monitor DKA admissions, ED visits, patient-reported burden, and health-economic outcomes.

Future Directions and Research Needs

Priority research areas identified by the panel:
– Prospective randomized trials comparing CKM-enabled care pathways to standard care with outcomes including DKA incidence, ED visits, hospitalizations, and patient-reported outcomes.
– Device validation in real-world conditions: accuracy across hydration states, pediatric and pregnancy physiology, and during rapid ketone rises.
– Implementation science: best models for remote monitoring, triage thresholds, and mitigating alarm fatigue.
– Cost-effectiveness and equity analyses to guide payer coverage and public health deployment.

Practical vignette

Sarah is a 28-year-old woman with longstanding type 1 diabetes on an insulin pump and a CKM device. After a weekend influenza-like illness with vomiting, her CKM trends from 0.4 mmol/L to 1.6 mmol/L over 4 hours. Her CKM issues an amber then red alert. She checks a POC blood BHB (1.7 mmol/L), measures her glucose (190 mg/dL), and follows her sick-day plan: maintains insulin delivery, takes small regular carbohydrate, increases fluids, and calls her diabetes team. The team advises intensified home monitoring and a repeat check in 2 hours. When ketones continue to rise, she is advised to seek urgent evaluation and is treated early, avoiding progression to full DKA. This vignette illustrates how CKM-informed trends, combined with structured action plans and access to clinical advice, can permit early intervention.

References

1. Dhatariya K, Bergenstal RM, Al-Sofiani M, et al. Continuous ketone monitoring for people with diabetes: international expert recommendations on the application of a new technology. Lancet Diabetes Endocrinol. 2026 Jan;14(1):82-92. doi:10.1016/S2213-8587(25)00331-6.
2. American Diabetes Association. Standards of Medical Care in Diabetes—2024. Diabetes Care. 2024;47(Suppl 1):S1–S300.
3. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic crises in adult patients with diabetes. Diabetes Care. 2009 Jul;32(7):1335–1343.
4. Wolfsdorf JI, Glaser N, Agus M, et al. ISPAD Clinical Practice Consensus Guidelines 2018: Diabetic ketoacidosis and the hyperglycemic hyperosmolar state. Pediatr Diabetes. 2018;19(Suppl 27):155–177.
5. U.S. Food and Drug Administration. FDA Drug Safety Communication: FDA warns that SGLT2 inhibitors for diabetes may result in a serious condition of too much acid in the blood (ketoacidosis). 2015. https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-warns-sglti-inhibitors-diabetes-may-result-serious-condition-too-much-acid

The consensus represents an important early framework for how CKM could be used to reduce the burden of DKA. Clinicians and health systems should adopt these recommendations thoughtfully, emphasizing patient education, device validation, and clear care pathways while awaiting prospective outcome data.